You are on page 1of 17

M.

Prasad Naidu
MSc Medical Biochemistry, Ph.D,.
It is an improvement over mean deviation

Measure of dispersion

Used most commonly in statistical analyses
Calculation of SD
First find the mean of series
Find the deviation or difference of individual
measurement from mean
Next find the sum of squares of deviation or
difference of individual measurements from their
mean
Now find the variance (Var) mean squared
deviation

var = (X - X)
2
/ n
If large sample size :




If sample less than 30 :
Square root of variance that gives SD
For ex :
Mean = 2+5+3+4+1/5 = 15/5 = 3

Var = 10/5 = 2

Uses of SD
It summarizes the deviation of a large distribution from
mean in one figure used as unit of variation
Indicates whether the variation of difference of an
individual from mean is by chance
Helps in finding the SE which determines whether the
difference between means of two similar samples is by
chance or real
Helps in finding the suitable size of sample for valid
conclusion.
The shape of curve will depend upon mean and SD of
which in turn depend upon the number and nature of
observation.

In normal curve :-
Area b/w 1 SD on either side of mean will include
approximately 68% of values in distribution
Area b/w 2 SD is 95%
Area b/w 3 SD is 99.7%
These limits on either side of mean are called
confidence limits
SD of normal curve
Standard normal curve
Smooth
bell shaped
Perfectly symmetrical
Based on infinity large number of observation
Total area of curve = 1
Mean = 0
SD = 1
Mean , median and mode all coincide

SD of normal curve
Bell shaped curve will show an inflexion on the
ascending as well as descending units of curve

If vertical lines are drawn from each of these points they
will intersect the X axis on either side of the mean at an
equal distance from it

A large portion of area under the normal curve has been
included in portion of curve b/w the 2 points of
inflexion

The distance b/w the mean and point of inflexion
either side is equal to SD and is denoted by a + sign
prefixed to it to indicate that it extends on either
side of mean.

If another vertical line is drawn an either side of
mean at a distance equal to twice SD most of
values in distribution table would have been
included in this part of curve

In most cases, + SD will include 2/3 of sample
values and mean + 2 SD will include 90% of
values.